Soft Classification of Satellite Data for Snow Mapping by using Multivariate Adaptive Regression Splines

2016-07-03
Kuter, Semih
Akyürek, Sevda Zuhal
Weber, William
Measurement of the areal extent of snow cover with high accuracy plays an important role in hydrological and climate modeling. Remotely-sensed data acquired by earth-observing satellites offer great advantages for timely monitoring of snow cover. However, the main obstacle is the trade-off between the temporal and spatial resolution of the satellite imageries. Soft or sub-pixel classification of low or moderate resolution satellite images is a preferred technique to overcome this problem. In this presentation, we represent fractional snow cover (FSC) mapping from Moderate Resolution Imaging Spectroradiometer (MODIS) data in Alps by using Multivariate Adaptive Regression Splines (MARS). The MARS model is trained in order to estimate FSC using MODIS surface reflectance data for the first seven reflective solar bands, Normalized Difference Snow Index and Normalized Difference Vegetation Index as predictor variables. FSC cover maps obtained by binary classification of higher spatial resolution Landsat ETM+ images are used for MARS model training and validation. The results of MARS FSC maps are also compared with the standard MODIS FSC product, and the results are given in terms of RMSE and coefficient of determination values.
Citation Formats
S. Kuter, S. Z. Akyürek, and W. Weber, “Soft Classification of Satellite Data for Snow Mapping by using Multivariate Adaptive Regression Splines,” presented at the 28th European Conference on Operational Research, 2016, Poznan, Poland, 2016, Accessed: 00, 2021. [Online]. Available: https://hdl.handle.net/11511/76585.